Association of toxin-producing Clostridioides difficile with piglet diarrhea and potential transmission to humans

Abstract

The pathogenicity of Clostridioides difficile in piglets remains controversial. It is unknown whether C. difficile control helps protect piglet health. To clarify the association between C. difficile presence and piglet diarrhea, isolates were obtained from piglets with and without diarrhea. In addition, to determine the genetic relationship of C. difficile from pigs and humans, we performed whole-genome sequencing (WGS) of C. difficile isolates. Diarrheal and non-diarrheal stool samples were collected from neonatal piglets from five farms in Japan in 2021. To clarify the relationship between C. difficile derived from pigs and those from human clinical cases, WGS of C. difficile isolates was performed. Toxin-positive C. difficile were significantly more prevalent in piglets with diarrhea, although the overall frequency of C. difficile did not differ between piglets with and without diarrhea. This observation indicates an association between toxin-positive C. difficile and diarrhea in piglets. However, further studies are needed to establish a direct causal relationship and to explore other contributing factors to diarrhea in piglets. WGS results showed that C. difficile sequence type (ST) 11 including the hypervirulent PCR ribotype 078 isolates derived from Japanese pigs were closely related to ST11 of overseas strains (human clinical and animal-derived) and a Japanese human clinical strain. Toxin-positive C. difficile may cause diarrhea in piglets and hypervirulent C. difficile are spreading among pigs and human populations worldwide.

Clostridioides difficile is a Gram-positive, anaerobic, spore-forming bacterium that causes antibiotic-associated diarrhea and pseudomembranous colitis in humans [36]. The Centers for Disease Control and Prevention consider C. difficile as the most important antimicrobial-resistant pathogen that threatens public health in the United States [9]. This pathogen secretes two major toxins, namely toxin A (TcdA) and toxin B (TcdB), which synergistically induce apoptosis in mucosal epithelial cells and cell-to-cell adhesion [41]. The toxins trigger an inflammatory cascade, leading to tissue damage and fluid extravasation [20]. While the in vivo pathogenicity of the third toxin, binary toxin (CDT), remains controversial, it is cytotoxic to intestinal cells in vitro [5]. CDT-producing strains have been linked to the development of more severe symptoms compared to strains that do not produce CDT [5].

While C. difficile is pathogenic to humans, the pathogenicity of C. difficile in animals remains controversial. The presence of C. difficile has been suspected to be associated with diarrhea in piglets [11, 18, 27, 33]. Diarrhea in piglets is an economically important disease owing to its high incidence, especially in suckling and weaning piglets. However, few studies have reported no association between C. difficile isolated from piglets and diarrhea in these animals [2]. It has not been established whether C. difficile control is required to protect piglet health. Thus, clarifying the association between C. difficile presence and diarrhea in piglets is necessary.

C. difficile is often classified via PCR ribotyping. Among the ribotypes (RTs), RT027 and RT078 are known as hypervirulent RTs [16]. RT027 and RT078 represent significant sources of severe C. difficile infections (CDIs), produce all toxins (TcdA, TcdB, and CDT), and exhibit high-transmission capacity [8, 40]. RT027 and RT078 are classified as sequence type (ST) 1 and ST11 in multilocus sequence typing (MLST), respectively. ST11 encompasses several RTs including not only RT078, but also RT126, RT127, and RT033. ST11 strains have been isolated from patients with CDI as well as from domestic animals worldwide [19, 23]. ST11 strains have been suggested to be associated with animals and patients with community-acquired CDI.

In Japan, although hypervirulent strains (RT027 and RT078) have been isolated from human clinical cases, their prevalence is low and severe cases are rare [38]. In contrast, C. difficile has been isolated in 0.8% of adult pigs [4] and 58% of piglets [45] screened in Japan, with RT078 isolated as the most important ST11 strain being detected in both cases. Notably, RT078 represents a major RT that causes human CDI in Europe [6]. Therefore, the relationship between C. difficile derived from pigs and those from human clinical isolates and overseas isolates should be clarified.

In this study, to clarify the association between C. difficile presence and diarrhea in piglets, C. difficile was isolated from piglets with and without diarrhea, and the isolates were characterized. In addition, to determine the genetic relationship of ST11 strains isolated from domestic animals and humans, we performed whole-genome sequencing (WGS) of C. difficile isolates.

MATERIALS AND METHODS

Sample collection and ethics

A total of 22 diarrheal and 34 non-diarrheal stool samples were collected from neonatal piglets (age, <21 d) from five farms in 2021 (Supplementary Table 1). The stool conditions were judged by commissioned veterinarians, and diarrhea cases were sampled without making fine distinctions based on the condition; all instances were simply classified as diarrhea.

Sampling was conducted in cooperation with commissioned veterinarian as a general health surveillance, and the study was not classified as an animal experiment. Informed consent was obtained from the farmers for the participation of their animals in this study.

Bacterial isolation and identification

Fecal samples were treated with alcohol for spore selection as described previously [4] and the isolated C. difficile was cultured on cycloserine-cefoxitin-mannitol agar (CCMA)-EX plates (Nissui Pharmaceutical, Tokyo, Japan) at 37°C for 36–48 hr under anaerobic conditions. Isolated colonies were purified by re-streaking onto CCMA-Ex plates followed by anaerobic incubation as indicated above. Up to three colonies suspected as C. difficile from each sample were sub-cultured for further analysis. The isolates were identified as C. difficile via matrix-assisted laser desorption-ionization time-of-flight mass spectrometry using the Bruker MALDI Biotyper system (Bruker Daltonics, Bremen, Germany) [12].

Toxin gene detection and toxinotyping

DNA was extracted using a commercial kit (InstaGene Matrix; BioRad, Hercules, CA, USA) according to the manufacturer’s instructions. The genes encoding toxins A, B, and CDT (tcdA, tcdB, and cdtA/B, respectively) were detected using multiplex PCR as described previously [31].

PCR-restriction fragment length polymorphism was used to amplify the A1, A2-A3, A3, and B1 regions of tcdA and tcdB in a 19 kb PaLoc sequence, and toxinotyping was performed according to a previous method [32].

PCR ribotyping

PCR ribotyping was performed as described previously [30, 43]. Briefly, the volume of the PCR mixture was downscaled from 50 μL to 15 μL, and the amplified PCR products were concentrated to a final volume of approximately 10 μL by heating at 75°C for 90–120 min. Electrophoresis was performed using 3% Metaphor agarose gels (Lonza Rockland Inc., Basel, Switzerland) at a constant voltage of 120 V for 4 hr to separate the PCR products. The PCR ribotyping banding patterns were analyzed using the BioNumerics program (Applied Maths, Sint-Martens-Latem, Belgium). Similarity and diversity were assessed by applying the Dice coefficient. Cluster analysis was performed using the Unweighted Pair Group Method with Arithmetic Mean algorithm. C. difficile strains 9689, 43593, 700057, BAA–1870, and BAA–1875 were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) to serve as reference strains.

WGS for RT078 and RT078-like strains

RT banding patterns of 60 isolates that matched or were similar to those of RT078 (control strain, ATCC BAA-1875) were analyzed, and nine isolates were found to vary among farms in WGS analysis. In addition, eight isolates of RT078 strains isolated from piglets and pigs in Japan in 2011 and 2012 were used for WGS analysis [4, 45]. Genomic DNA of these strains was extracted using the Qiaquick PCR Purification Kit according to the manufacturer’s instructions (QIAGEN, Hilden, Germany). The contig was mapped with 300 bp paired-end reads obtained using Nextera XT and MiSeq sequencing (Illumina, San Diego, CA, USA).

WGS analysis of ST11 strains including RT078, RT126, and RT127

The WGS results of ST11 strains (RT078, RT126, RT127, and RT033) were downloaded from public databases [3, 23, 28]. Subsequent comparative analysis was performed using the sequences of 17 strains sequenced in this study and 184 downloaded sequences of ST11 strains (Supplementary Table 2).

The quality of sequence reads was checked using fastqc in the Galaxy platform (Galaxy ver. 0.72) [1]. In addition, sequence reads of tested strains were mapped to the reference genome M120 (GenBank accession No. NC_017174) and variants of the core genome were determined using Snippy (v0.2.0.) [37]. Subsequently, several results of Snippy were integrated into core single-nucleotide polymorphism (SNP) alignment using Snippy-core. These results were used to calculate pairwise core genome single-nucleotide variant differences between isolates and to generate maximum-likelihood phylogenies. Phylogenetic trees were generated using RAxML (v8.2.10) with a generalized time-reversible model of evolution and CAT approximation of rate heterogeneity and were curated using iToL v4 [26, 42].

MLST and antimicrobial resistance gene analysis for RT078 and RT078-like strains

MLST analysis was performed according to the PubMLST protocol and database (https://pubmlst.org/organisms/clostridioides-difficile). The antimicrobial resistance genes were detected using ResFinder v4.1 with default parameters on the CGE server (http://www.genomicepidemiology.org).

Statistical analysis

Statistical significance was determined using a χ² test using a cutoff P value of <0.05.

RESULTS

Isolation of C. difficile from piglets with or without diarrhea

C. difficile was isolated from 73.7% of piglets with diarrhea (16/22) and from 55.9% of piglets without diarrhea (19/34; Table 1). In total, 46 and 57 strains of C. difficile were isolated from piglets with and without diarrhea, respectively (Supplementary Table 3). The frequency of toxin-positive strains (harboring tcdA and tcdB) isolated from piglets with diarrhea (63.6%; 14/22) was significantly higher than that in piglets without diarrhea (50%; 17/34; P<0.05). The prevalence of ST11, which is classified as a highly virulent strain, among piglets with diarrhea (63.6%; 14/22) was significantly higher than that in piglets without diarrhea (32.4%; 11/34) (P<0.05). The rate of RT078 isolation, which is classified as one of the ST11 strains, was significantly higher in piglets with diarrhea (45.5%; 10/22) than in those without diarrhea (17.6%; 6/34) (P<0.05).

Table 1. Relationship between possession of Clostridioides difficile and piglet diarrhea.

	Diarrhea (n=22)	Non-diarrhea (n=34)	P value
C. difficile positive	16 (73.7%)	19 (55.9%)	0.09
Toxin-positive C. difficile	14 (63.6%)	17 (50.0%)	0.04*
ST11 C. difficile	14 (63.6%)	11 (32.4%)	0.01*
Ribotype 078 C. difficile	10 (45.5%)	6 (17.6%)	0.03*

*P<0.05.

PCR ribotyping and toxinotyping

In total, 103 isolates were classified into seven RTs (DP1 to DP7) (Fig. 1). In few cases, multiple RTs were obtained from a single farm (Farm C, D, and E), while only certain RTs were obtained from other farms. RT078 (DP1) and similar RTs (DP2 and DP5) were observed in 60 isolates. Among these 60 isolates, 40 were classified into toxinotype V, and the remaining 20 isolates were classified into toxinotype XXVIII.

Fig. 1.

PCR ribotype profiles of Clostridioides difficile isolated from piglets. In total, 103 isolates were obtained from piglets in this study. ATCC strains were used as control strains.

MLST and antimicrobial resistance genes

Nine selected isolates of RT078 and RT078-like types obtained in 2021 were classified as ST11 via WGS analysis (Table 2). In addition, eight RT078 isolates from pigs obtained in 2011–2012 were classified as ST11 via WGS analysis. From the results of WGS, toxinotype XXVIII strains (RT DP2) in this study were classified as RT127.

Table 2. Characterization of ST11 Clostridioides difficile isolated from pigs and humans in Japan.

Thirteen of the 17 ST11 C. difficile isolates from pigs harbored both tetracycline and aminoglycoside resistance genes (Table 2). The remaining four isolates did not harbor any resistance genes. In addition, one of the seven human-derived ST11 C. difficile strains previously isolated in Japan harbored aminoglycoside resistance genes. In contrast, the remaining six isolates did not harbor any resistance genes based on Resfinder analysis results.

Comparison of WGS results of ST11 C. difficile

The results of core-genome SNP analysis of ST11 strains isolated in this study (17 isolates) for which WGS was performed along with analysis of isolates from domestic and foreign sources (152 from human clinical cases and 32 from livestock) are shown in Supplementary Fig. 1. These ST11 isolates were classified into six groups (groups A to F) based on SNP analysis in this study. Fourteen toxinotype V isolates derived from piglets in this study belonged to group A (total, 128 isolates). Group A isolates belonged to RT078 or RT126. Additionally, three toxinotype XXVIII isolates derived from piglets in this study belonged to group D (total, 19 isolates). In group D, all isolates belonged to RT127.

Further analysis was performed to clarify the genomic similarity among group A isolates (Fig. 2). C. difficile strains derived from Japanese pigs were similar based on year and farm of isolation. Further, they were not clearly distinguishable from FD267 derived from human clinical cases in Japan and from strains isolated overseas.

Fig. 2.

Phylogenetic tree of group A comprising ST11 strains (n=128) based on evolution of core genome single-nucleotide polymorphisms. Red background indicates strains isolated in Japan. The blue circle indicates a human clinical case in Japan.

Further analysis was performed to clarify the genomic similarity among group D isolates (Fig. 3). The genomes of the 19 isolates belonging to Group D were highly similar; Group D also included three RT127 isolates from Japanese pigs and six RT127 isolates from clinical cases in Japan.

Fig. 3.

Phylogenetic tree of group D comprising ST11 strains (n=19) based on evolution of core genome single-nucleotide polymorphisms. The background color indicates the country of isolation. A circle next to the strain name indicates whether it is of human clinical case or pig origin. M120 is used as reference strain.

DISCUSSION

The frequency of toxin-positive C. difficile was significantly higher in piglets with diarrhea than in piglets without diarrhea, although the overall frequency of C. difficile did not differ between piglets with and without diarrhea. C. difficile is frequently isolated from the feces of young mammals [7], and a certain proportion was isolated from piglets without diarrhea in this study as well. However, the frequency of ST11 strains including RT078, which exhibit high toxin production [10, 18], was relatively greater in piglets with diarrhea in this study. ST11 was also isolated from non-diarrheal piglets; however, there is a variance in toxicity even within the same ST [14]. Toxins produced by C. difficile affect the intestinal tract of pigs [46]. These findings indicate an association between the presence of toxin-producing C. difficile and the occurrence of diarrhea in piglets. However, these results should be interpreted with caution as they represent circumstantial evidence; while they suggest a potential link, they do not establish a direct causal relationship. The causes of diarrhea in piglets are not limited to C. difficile alone (other pathogens and environmental factors are also involved). Further research is required to determine the precise role of toxin-producing C. difficile in the etiology of diarrhea in piglets.

Core genome SNP analysis showed that C. difficile strains with similar genotypes were spread across the farms. C. difficile is a spore-forming bacterium that shows high resistance to disinfection and cleaning agents [15, 34]; therefore, a specific genotype of C. difficile may persist in the farm environment. These results suggest that piglets primarily acquire C. difficile from the farm environment. In hospitals, the use of environmental disinfectants such as sodium hypochlorite is effective in controlling C. difficile [35]. Toxin-producing C. difficile can cause diarrhea in piglets, indicating the importance of effective disinfection for controlling C. difficile spread in the farm environment to prevent diarrhea in piglets.

Most of the RT078 strains derived from piglets harbored both tetracycline and aminoglycoside resistance genes. In contrast, isolates classified as RT127 in the same ST11 harbored no antimicrobial resistance genes. The possession of antimicrobial resistance genes differed between RT078 and RT127; the exact reason for this is unclear and requires further study. However, the frequency of tetracycline resistance genes in RT078 derived from pigs may reflect the status of tetracycline usage in pig farms in Japan [13, 29].

RT078 derived from Japanese piglets were closely related to overseas strains (human clinical and animal-derived) and a Japanese human clinical strain (FD267). It has been reported that clonal strains of RT078 are spreading internationally [25]. In addition, RT127 C. difficile isolated from Japanese piglets were closely related to RT127 isolates obtained from overseas and human clinical cases in Japan. Previous studies have suggested that RT078 was introduced into Japan via sows [45]. RT127 is also one of the three toxin-producing strains classified as ST11, along with RT078, and it has been isolated from livestock and patients with CDI in Australia, Japan, and Taiwan [24, 44]. Keessen et al. reported that the C. difficile carriage rates in farmers was high [21]. Further, the genetic characteristics of C. difficile derived from farmers and pigs were similar [22, 28, 39]. These results provide evidence that ST11 C. difficile found in pigs may have been introduced into Japan from overseas via sows and appear to have undergone independent evolution on farms before potentially spreading to farmers and other humans. Given the insufficient epidemiological information, these conclusions are tentative and require confirmation through additional epidemiological studies.

In conclusion, reducing the contamination of C. difficile spores in the swine farm environment is important not only to reduce the risk of infection from livestock to humans, but also to improve the health of piglets. Although the number of clinically reported cases of CDI in Japan is limited compared to that in other countries, the identification of genomic similarities between human and porcine strains of ST11 supports the public health risk of ST11. Recently, RTs from ST11 that are more virulent than RT078 have been identified [17]. In the future, surveillance and control measures for C. difficile (especially ST11) in the livestock environment should be implemented to control the spread of C. difficile and its transmission to the human community via livestock feces.

CONFLICT OF INTEREST

The authors do not have any conflicts of interest to declare.